Natural Language Processing · Trustworthy AI
Teaching machines to tell authentic text from the synthetic.
I'm a first-year PhD student in Computer Science at Texas State University, working in the HILL-NLP lab under Dr. Veronica Perez-Rosas. My research sits at the intersection of language models and trust: detecting machine-generated content, studying deception, and building human-centered AI that people can rely on.
A short bio
I recently completed my Master of Science in Computer Science at Texas State, where my thesis — SAFE: Suspicious vs. Authentic Feedback Evaluation for Detecting LLM-Generated Reviews on E-Commerce Platforms — examined how well transformer classifiers can separate genuine customer reviews from ones produced by large language models. That work has grown into the questions I now pursue as a doctoral researcher: provenance, robustness, and the limits of automated detection.
Before graduate school I earned a Bachelor of Science in Computer Science from the National University of Computer and Emerging Sciences (FAST) in Pakistan, and spent several years as a software engineer building trading systems and distributed backends. That engineering background shapes how I approach research: I care about methods that hold up outside a benchmark.
Currently
Preparing SAFE for a workshop venue
Extending the thesis into a publishable study on LLM-generated review detection, while defining a broader dissertation direction in machine-generated text provenance and adversarial robustness.
What I work on
- Detection. Distinguishing human from machine-generated text across domains, prompting strategies, and model families.
- Robustness. Understanding when detectors fail under paraphrase, persona conditioning, and adversarial pressure.
- Trust & deception. Studying deceptive language and how NLP can support, rather than replace, human judgment.
- Translation. Machine translation for NLP — multilingual modeling and translation quality.
Explore my research, teaching, and service — or reach me at har98@txstate.edu.